Still Believe in 200 Rank Factors & Santa Claus?

[martinibuster]

Mod's note: original title and description lines...
Still Believe in 200 Rank Factors & Santa Claus?
Search has Changed. So Why Do SEOs Talk About This?

The whole idea of 200 factors that Google checks a site for is quaint, like Jell-O . A checklist of things you have to do right is convenient but that's not how search engines work.

The 200 ranking factors has been with us since about a year or two after Google was out of diapers. There have literally been thousands of changes to the Google algorithm.

Do you still believe there are 200 ranking factors?

[edited by: Robert_Charlton at 6:49 am (utc) on Oct 29, 2016]
[edit reason] Description line not used in this forum. Put it into the post. [/edit]

[waynne]

There are still lots of ranking factors which compound each other only a few bare a major impact in the SERPS. 200? Probably more than that.

There are also now deranking factors where if a site looks optimised the optimisation is ignored and if it's really over optimised it may get penalised and derank.

[EditorialGuy]

I don't think 200+ ranking factors were ever a useful "checklist" for site owners and SEOs, if only because the details of Google's algorithm (including the weighting of different factors and how they interact) have never been made public.

[aristotle]

Google can adjust the weightings of these factors to bias the results in various ways.

An obvious example is relevance. One long-standing problem for Google is that spammers tend to create pages that are highly relevant for certain search terms. Back in the days when relevance was a bigger factor, these spam pages would often rank high in the results.

Some years ago Google realized that boosting the rankings of big brands and big organizations would help to keep spam out of the top results. But in order to boost the rankings of big brands and big organizations, they had to reduce the weight given to relevance. This is why we now see so many less relevant pages from big brands ranking above more relevant pages from smaller sites.

The point is that Google can adjust the weightings of various factors to get results that are more in line with what they want.

[martinibuster]

Some years ago Google realized that boosting the rankings of big brands...

I thoroughly debunked it in an article titled, Brand is Not a Ranking Factor [thesempost.com]. :)

It doesn't help matters when some people lazily theorize about ranking factors without providing any supporting citations to research or patents. And this circles back to the 200 ranking factors, which includes stale and useless notions as keyword density. keywords in URL, etcetera. Leaving aside how stale and utterly useless some of these commonly cited factors are, simply the idea of 200 factors is simply put, naïve. The reality is closer to thousands.

[iamlost]

Search has Changed. So Why Do SEOs Talk About This [ 200 Rank Factors ]?

Or: Search has Changed. Why do SEOs talk about 90%+ of what they talk about?
Because 90%+ of SEOs are either:
* parrots who never forget a marketing point.
* tool makers with products not fit for purpose.

From the very beginning I made two points:
1. that knowing how many 'factors' didn't actually tell us anything the slightest bit useful such as (1) what the factors actually ARE, (2) what the inputs for each actually ARE, or (3) the weightings and sequence of each.
and the most mind blowing point of all:
2. even if Google (or any other SE) printed out it's entire algorithm and posted it for all to enjoy...
* few webdevs/SEOs would actually have a clue, in a practical sense, what they were looking at or how to take advantage of the knowledge;
* that if on doing so, Google (or any other SE) subsequently altered this that or the other even in the slightest bit the results for edge cases would be significantly affected; just as with every Google update.

And I was told by all and sundry that I didn't know what I was talking about. Well, folks, I'm still in business still ranking well for oodles of converting Google (and other SE) traffic while a good many of the nay sayers are not.

There are SEOs and webdevs to whom I pay the closest of attention and give the greatest of respect but they tend to be the quietest not the most popular. But then iamlost.

After Note:
Google doesn't actually have a problem with MFA spammers, it has a problem with getting called out about MFA spammers. Those critters have and do make not only a great deal for themselves but also for Google who take their house percentage right off the top. The 'rules' change by vertical for a very solid bottom line reason.

When the result is that the spammers influence the non-spammers sufficient to pull the overall results too low they make a change; when some legal challenge impacts common behaviour they make a change; etc. By just enough percentage to look significant without actually impacting the bottom line as much.

The last great 'con' they got called on was impressions. It wasn't something unknown or new just that the accusers were - finally - the folks paying the bills and the conmen SEs and ad agencies had to respond, which they did as minimally as possible: 50% of an ad visible for 1-second, really?!

Bleh.
A pox on all their houses.

[EditorialGuy]

Leaving aside how stale and utterly useless some of these commonly cited factors are, simply the idea of 200 factors is simply put, naïve. The reality is closer to thousands.

Well, Google has been saying "more than 200 ranking factors" for quite some time. (Italics mine.) :-)

[vordmeister]

26 of them was enough for me.

[NickMNS]

It is fact that Google use machine learning as part of their search algorithm. Therefore Google uses far more than 200 features for ranking, and not even Google know what of those are.

[tangor]

I looked. Maybe not that hard .... Susan (Miracle on 34th Street) "I believe... I believe..." Santa Claus exists, so do 200 points of g. It's a fact.

I like g traffic, don't get me wrong, but I have never coded for g, the 200 or not!

Disclaimer: I do not depend on adsense or adwords. My sites are what they are, either info or commercial and all I want from g is the signage (listing) on "main street" (serps) and I do that by having good sites/content. Has also worked well on the other "main streets" (bing, yahoo, duckduckgo, etc)

One CAN code to an algorithm or one can code for business/work and go from there. What one NEEDS to do is not code stupid to game an algo for position, else one ends up in the barnyard with all the animals.

[doc_z]

Yes, search has Changed. In general systems are getting more complicated with time - the same applies to Google.

There are a lot of signals like semantic, social media, reputation... which might became a ranking factor since the early days. Therefore, I expect that the ranking factors increanse over the years.

[ogletree]

I remember Matt Cutts talking about how they constantly adjust the dials of a lot of factors. I think he even spoke about 300 factors and that it was actually a lot more than that because many of the factors had sub factors and that there were more than one algo. This is from stuff he said a long time ago and could be all wrong now.

If you go through and read all the old posts and lists of factors that people have on their blogs one thing you will notice is that they are all considered very very minor factors.

If you create a good website with content that people actually read and find useful and don't do any negative on page SEO you will rise to the top and will get lots of organic traffic.

If your trying to start a business with only SEO in mind and you do nothing else to get traffic your going to have an uphill battle.

[mack]

I think it's safe to assume there are way more than 200 ranking factors, but up until now we are focusing in these factors as part of a "one size fits all" algo. It may be a case that some factors are conditional. Some criteria has to be met before a page will be subjected to another factor. page A may be subjected to a lot more ranking factors than page B.

Mack.

[martinibuster]

Certainly, ranking factors still exist. Links are an obvious ranking factor. There are also signals specific to local and mobile.

But there are also considerations of how likely something is to satisfy the user intent. How do you even put that into a box and tie a bow around it and call it a ranking factor?

[ogletree]

200 Ranking factors is what keeps SEO companies in business. Very few SEO's know how to get links so they have to charge for something. The vast majority of money spent on SEO is doing the below 1% stuff. If your paying under $1000 a month your spending most of your money on the little stuff and sometimes on nothing at all. I have even seen SEO companies charge more than $1000 a month for the little stuff and even messing that up.

Every single company that does website design claim to be SEO's. Everybody claims to be an SEO now days. You talk to these people and they don't know anything about SEO. Not even the most basic stuff. I can't tell you how many sites I have seen where an SEO just came in and added tons of keywords to the content. The pages were horrible and made no sense to the reader.

I have tried to get out of the SEO business but I can't keep my mouth shut when I see my PPC clients getting fleeced by disreputable SEO companies.

[smilie]

>>martinibuster: simply the idea of 200 factors is simply put, naïve. The reality is closer to thousands.

If you run a dataset of billion of ...pages...against a set of rules...you may produce thousands ..if not hundreds of thousands ranking points. Each could be different for pages vs. sites too. Technically there's no limit to how many factors.

But the cold reality is that one needs to worry only about a few simple ones on + side and several negative ones.

Some years ago Google realized that boosting the rankings of big brands...

>>martinibuster: I thoroughly debunked it in an article

I don't see any debunking there. I also see Amazon and Walmart dominated serps to a point where Amazon is now close to 30% ALL US ONLINE COMMERCE (per Internet Retailer). You were saying what about "debunked"?

[smilie]

>> Every single company that does website design claim to be SEO's. Everybody claims to be an SEO now days. You talk to these people and they don't know anything about SEO. Not even the most basic stuff.

On one of my projects I've hired an SEO firm with very specific list of objectives, all white hat and true marketing related. Fired them in two months breaking contract, for breach of contract on their end of feeding me useless SEO-101 type junk. And that was a nationally recognized firm.

As tens of thousands of small businesses drop dead like flies based on Google's Death Grip, they go to any magic that may be left out there. These SEO companies pray on it, feeding them BS (talking the talk) and delivering a fairly poor quality SEO-101 type of stuff 80% of which isn't working (not walking the walk). And if you need advanced stuff there are companies out there willing to pay thousands per month for it, and they are out of dying small businesses reach. Talking about misallocation of capital.

[martinibuster]

I also see Amazon and Walmart dominated serps

Links is an important factor why Amazon wins. What makes Amazon large is also their weakness. I beat Amazon and Wal-Mart for my own rankings (on informational sites!) and for e-commerce clients.

Because there is no patent or scientific research to show that a Google Brand Bias is something Google consciously builds into the algorithm, one can only conclude that the idea of a Google brand bias is little more than a crutch for explaining why other sites rank better.

There are citations to a speech by Eric Schmidt to explain the Google Bias but I debunked that by showing the exact circumstance and context of his words to show that it is inappropriate to use those words to propagandize a false notion of a Google brand bias.

I debunked it with citations. If any of you still believe in a brand bias then step up with citations to Google patents and research papers outlining how they intend to promote brands regardless of quality or user experience. I would also love to see your citations to patents and research that outline what brand signals are.

Until you or anyone can produce the proper scientific citations, the concept of Brand Bias is an unfounded opinion, like flat earth theories. These false notions are dangerous to believe because bad information is never helpful. I say this to help others: One must understand the difference between facts and excuses.

As far as advanced information being out of reach that's really a case of not knowing where to look. There are many paid resources that offer various level of information.

[edited by: martinibuster at 5:52 pm (utc) on Oct 31, 2016]

[smilie]

martini, I see tens of merchants around me dying and going out of business. While every time I search for some product Amazon is in top 5. That's enough evidence, you don't need citation.

>> Until you or anyone can produce the proper scientific citations, the concept of Brand Bias is an unfounded opinion, like flat earth theorists.

You still think it is "scientific" ? It is more of a legal guideline.

Or do you want me to copy a page from a secret publication that is a guiding document for several companies that hire thousands of page rankers for big SEs? Where "a page on a brand website" is a significant positive ranking factor?

A rabbit hole goes pretty deep

[martinibuster]

you don't need citation.

You absolutely need a citation. Without a citation it's just an opinion with no weight on reality.

Google is not a black box. Everything that happens at Google is in a patent or in scientific research papers. All of it. If you can't find a proper citation then it's just something someone made up, a mere opinion.

There's value in opinion that has an actual foundation.

a secret publication

Download Google's Quality Ratings Guidelines. Half the algorithm's hiding in plain sight between the lines. ;)

[NickMNS]

Google is not a black box. Everything that happens at Google is in a patent or in scientific research papers. All of it. If you can't find a proper citation then it's just something someone made up, a mere opinion.

Not everything in Google's patents is implemented, or implemented as described, and nothing prevents Google from implementing things that are not patented or published.

The absence of evidence does not constitute proof. That is not how the scientific process works.

Your claim is as much opinion as Smilie's.

[ogletree]

It is possible for a factor to not have a direct citation. It could be the result of a side affect of other citations combined. Sometimes their are temp or long term factors that Google did not intend and remove at some point maybe quickly maybe it takes 5 years or they have not figured it out yet. I have met SEO"s that have their list of tricks they don't tell anybody because they did extensive testing.

I find that when an SEO gives away their trick it is because it no longer works or not as well as it used to. I have had tricks I used in the past that no longer work but worked very very well for a time.

[martinibuster]

Not everything in Google's patents is implemented,

I agree.

or implemented as described

I also agree.

and nothing prevents Google from implementing things that are not patented or published.

Again, we are in agreement. :)

Google and other search engines license technology from universities, too. Which is why I say it's a huge mistake to solely focus on Google patents and Google research papers. There is other research going on worldwide that could be used at the search engines.

It is in the business interest of everyone in the search marketing industry to have some background information on Information Retrieval. Even just a little. How can one proclaim proficiency in influencing the search engines without knowing even the basics of how search engines function?

It was only last year that members on WebmasterWorld were still clinging to keywords in the H1 tags (some still do). How is it that we know that these things aren't as important anymore? Well, there is observation. That's one leg to stand on. Then there is knowledge of things like Hummingbird. That's another leg to stand on. The more information you attain the clearer you are able to see the truth of what is actually happening.

The absence of evidence does not constitute proof. That is not how the scientific process works.

Science is evidence based. That is how science works. Nick, are you opposed to examining a theory and testing it to see how right it is? Especially when there is money involved?

So all I am asking is, let's set aside the phrases taken out of context that were used to prop up the theory of Brand Bias and take a look at it rationally and logically. That's how science works. That is the essence of scientific inquiry.

Evidence has been presented that Google prefers brands in order to explain why Amazon/Wal-Mart ranks. But this evidence consists of words taken out of context. The theory is without foundation.

So let's take a look at some other evidence. The smoking gun can always be found in patents and research. But the thing is, it does not exist. There is no scientific smoking gun. Not from Google, not from a university. Not from Baidu and not from Bing. The evidence could be a research paper out of Stanford University that shows how brands are important to satisfying consumers and then lays out signals of brands.

But of course, that defies logic. Why go through the trouble of creating an algorithm that expressly favors brands when you can simply add a weighting to Amazon, Wal-Mart and Wikipedia.

This is business, not a religion. Basing business decisions on myths is illogical and bad business. All I am asking is, show the proof. Pointing at the SERPs and saying it's Brand Bias is as intellectually lazy as pointing at the moon's craters and saying it's made of swiss cheese.

[edited by: martinibuster at 8:15 pm (utc) on Oct 31, 2016]

[tangor]

Within "science" is "empirical" which is solely based on observable evidence (it is what it is) and one also has to take into consideration that a patent can be abused. After all, humans are running the show, and we all know how trustworthy they can be--- especially when large sums of filthy lucre and power are involved.

[martinibuster]

one also has to take into consideration that a patent can be abused.

Please elaborate. I'm not sure what you mean by that. Do you have an example of what an "abused" algorithm looks like?

I can certainly give you an example of a biased algorithm. Yahoo's trust rank algorithm was biased against small link communities. This in fact was discovered and a new algorithm was proposed by another group of scientists who proposed a topical trust rank, one that would not be biased against small sites and small communities of sites.

We can't ever be ahead of Google's curve, but we can be up to date. It's possible. You can't afford to sit around waiting for Google to announce things because by then they've been eating your SEO lunch for eight years already. Most importantly, it does not make business sense to make decisions based on information that is wholly without foundation.

[edited by: martinibuster at 9:00 pm (utc) on Oct 31, 2016]

[smilie]

>> martini: Google is not a black box. Everything that happens at Google is in a patent or in scientific research papers

This is a naive, "we are a tech geeks company" position.

Here's another position: Google is not a tech geeks company. It is a defense contractor with giant DARPA and classified technology ties. That is also owned by banks and stock market , and therefore, needs to show profit growth every quarter.

Once you understand this, you will get that , besides tech and patents, there are multiple levels. There will be , besides tech and open information, closed information, For Internal Use Only guidelines, For Our Partners only guidelines, legal, market-related guidelines , how-we-make-more-money-every-quarter guidelines etc. etc. There's several layers.

So. As Google failed to address SERPs quality with science, as there is none that is capable of identifying even stupid paid link from a non-paid link. To address this problem that they are covering up, Google (and Miscrosoft as well as other big search engines like Baidu) employ 3-rd party companies that evaluate SERPs quality and clean what they don't like. And those guidelines are written for thousands of searchers to evaluate a page very quickly (under a minute) and possibly hundreds of pages in one work period. So they are simplistic to the core , due to difficulty to have anything scientific here and due to employing average off the street and/or young, often non-tech evaluators.

And suffice me to say that I've seen one document, a guideline for these searchers. And there in plain site it says "page from a brand gets a boost" - I am probably shorting a 50-100 page document into one sentence, and there's lots of other ifs , maybes and buts. But this document is out there and it is affecting SERPs with strong bias toward Brands.

<<== this is the actual reality behind "we are totally only science" facade.

[tangor]

Abused patent? How about EpiPen? :)

In the search engines (and I include all of them) the "patent" is how it gets done, but it is humans who decide what that patent gets to do ... and there is your bias. Desperately trying to keep "politics" out of this forum, but that is becoming more difficult every day as even MSM (which loves g) is reporting linkages, agreements, people, money, backroom deals, lobbying, ex-googler's in government, and all kinds of icky things these days.

The empirical evidence of manipulation is that some sites get a boost because of big brand while others with better content for the user are deflated or ignored. This we can see with our own eyes, day after day.

I believe in the "200" factors ... I just believe they have been modified to skew one direction over another .... and that is abusing a patent (only providing it x info instead of a-z info) then pointing at the algo and saying "we can't control what it finds, it's all machine code!"

I don't depend on g. Yes, I like whatever traffic I get, but years ago I ceased to "groom" sites for g and instead returned to the real job of presenting for the USER, not a company which has moved into silos (killing thousands of web sites), uses content on their front page lifted from websites without compensation (or even a really visible link) and co-opted the internet advertising machinery with a bait and switch (early on make millionaires of a few until all are hooked, then change the dial so WE get rich).

I believe there are 200 or many more factors. I just don't think they are geared to level the playing field, allow competition, or even provide the best user results. And any webmaster who tries to code to "the factors" is cruisin' for a bruisin'.

Excuse my rant, kiddies.

[NickMNS]

@Roger, just to set the record straight, I tend to agree with your point of view regarding Google's handling of brands. Some big brands have very strong association with certain product, Kleenex is great example of this. It is only normal that if you search for facial tissues that Kleenex brand will come up in search. How Google comes to make this association is some what moot. But the association exists in the search results, whether it is baked into the algo or not is of course the subject of this debate. But again in my view this is moot.

This association exists with shoppers at the supermarket. Put a soft drink on the market with an unknown brand, then get it on the shelf next to Coke or Pepsi and see how well it performs. There is no search algorithm at play, the branded product outperforms. That is why brands have value. It is only natural to expect a search algorithm to mimic this behavior.

Science is evidence based. That is how science works. Nick, are you opposed to examining a theory and testing it to see how right it is? Especially when there is money involved?

Read what I wrote carefully, I wrote the absence of evidence does not constitute proof.

I am definitely opposed to testing theory determine their degree of rightness. A theory should be falsifiable, that is evidence shows that it is false, or it is not false.

Here is your theory as I read it "Ranking factors or parameters used by Google in their search algorithm have been published in the form of a patent or research paper at some point". Then given that this theory is true one can infer that since there is no reference to a brand bias, brand bias does not exist in the algo. The problem is that your theory is not testable, for it to be testable you must be able to show that every single ranking factor or parameter used by Google has been published in the form of a patent or research paper, and that only these factors are in use. You cannot do this. It is impossible for you or anybody else outside the inner circles of Google. In fact it may be impossible for anybody to prove it at all. We can't even decide on how many such factors exist.

I agree with Smilie in so much as you have not debunked anything, all you have done is provide another opinion.

@Smilie you lost me at "Here's another position". The Google is evil argument, really? Google is evil therefore they bias towards big brands.

[martinibuster]

But the association exists in the search results...

We totally agree. Nick, you are almost there and now I'm going to push you in the right direction. :)

The association exists. But it's not a brand bias. Brand bias is the lazy anti-intellectual conclusion, like the blind guy holding an elephant's trunk and declaring that an elephant resembles a snake.

Google is indeed biased. And these are the clues to Google's bias:

• The answer is written throughout the entire Google Quality Raters Guidelines.
  
  
• The answer is in the first lines of Google's policy statement [google.com] of the ten things they hold true.
  
  
• The answer is in Google's various (and misunderstood) patents for studying click logs.

Google aspires to please their users.
They want their users to come away satisfied. Your local supermarket aspires to please it's shoppers. Walk down the cereal aisle. What do you see?

• Fruit Loops
  
• Captain Crunch
  
• Lucky Charms

Google's algorithm is tuned to please users. What do you think all that click log mining is all about if not to find out what pleases users? Lastly, what do you think happens when your focus turns away from showing people the correct answer to a search query to showing people what will make them happy?

IF you read the patents about clicks logs, click backs, CTR and dwell time and machine learning the focus is always on pleasing the user. Do the metrics show that the user was happy? If yes then the search engine succeeded. But is showing people what they (apparently) want to see the best response? Google is indeed biased and it is biased toward showing people what pleases them.

I did a search related to gluten and the first response is logical, from a reputable site. The second position is by some site I never heard of that features an article about ways gluten can hurt me. Why is that site there? Because people like to eat fruit loops and supermarkets tend to sell people what they want to eat and Google's search engine is tuned by design to please their users.

This isn't something I just wrote on the fly. These are insights I've had for the past few years and in fact I've been wanting to express in an article for publication. The tentative title has for the past few years been The Fruit Loops Algorithm.

Over two hundred ranking factors?
It's 2016, no SEO should be talking about 200+ ranking factors. That's a joke. The science of information retrieval is larger than two hundred outdated ranking factors.

[Shaddows]

Going further on MB's point.

Your local supermarket isn't remotely interested in being fair to their suppliers. Indeed, Walmart in the States and Tesco in the UK are well known for squeezing the life out of their vendors. For "vendor" read "website".

Your local supermarket isn't interested is giving the customer what's best for them. Nutritious, wholegrain cereal or Fruit Loops? The muesli supplier can shout til they're blue in the face that the muesli is much higher quality than the Fruit Loops. That if only consumers were given the chance to see the wholesome fare, they would be much more satisfied. That filling the cereal isle with Fruit Loops amounts to a conspiracy against both the consumer and the muesli vendor. That UberMarket^(tm) has an inherent brand-biased agenda in pushing Fruit Loops vs muesli.

UberMarket is not even that interested in satisfying customers as a good in itself. Customer satisfaction is what drives profits. If you work against your customers' needs, it causes them to go and check out AlterMarket down the road. Exploiting your customers for a fast buck today is a sure fire way to not having them there tomorrow. It's a much better strategy to keep them satisfied, and take a cut. Fast-buck thinking does NOT give quarter-on-quarter, year-on-year growth. If gives you a big spike followed by a horrific crash.

As yet, I have not noticed Google crashing.